Forged Alliance Forever Forged Alliance Forever Forums 2015-01-31T00:40:43+02:00 /feed.php?f=42&t=9326 2015-01-31T00:40:43+02:00 2015-01-31T00:40:43+02:00 /viewtopic.php?t=9326&p=92603#p92603 <![CDATA[Re: Suggestion for next patch]]>
Gorton wrote:
Remember that there are tens of thousands of people who play - and a very small amount have complaints. It's not as widespread as you think.


Worth double posting to add: I work in tech support for a huge software company, and one of the principles of it is that if one person complains, 10% of your end users have experienced the problem.

Don't mistake lack of complaint for lack of issue. Especially when, I as I said in my post just above, reporting and following up on problems is so un-intuitive.

Though I have a hard time believing not a single player exists that has never experienced a bug or a crash, I acknowledge there are people who haven't noticed or don't care about the bugs they experienced. That being said, I think you are underestimating the impact if you think the only ones experiencing the problem are the ones that complain about it.

Out of curiosity, what percentage of people with a FAF account have an account on the forums? And what percentage of forum users are active (post, say, more than once a month)?

Statistics: Posted by codepants — 31 Jan 2015, 00:40


]]>
2015-01-30T04:35:58+02:00 2015-01-30T04:35:58+02:00 /viewtopic.php?t=9326&p=92566#p92566 <![CDATA[Re: Suggestion for next patch]]>
Gorton wrote:
I don't think you understand me.

(...)


Actually, Icedreamer's response is more what I was looking for (thanks!). But in response to this, I want to start by saying thank you, I appreciate all that you do. I know it is your free time, I know you have precious little of it, and I really do appreciate every ounce of effort you put in, even if it does make a new bug every now and then.

That's exactly why I'm probing -- I want to make sure the game is as good as it can possibly be. I want to make sure your work doesn't go to waste.

Second, I agree that a bug report system would be *so* nice. I want to report bugs/crashes/etc, but I honestly have seen a number of threads and a number of different ways to do it, none of which were both intuitive or easy sounding ("do such and such setting to make sure when FAF crashes it creates a log..."). IMO the best way would either be (A) another tab at the bottom of FAF called "Report Bug," with a window for a description, some radio button questions ("Was this a... crash? hang? etc.), and a button to auto-attach your latest replay, crash script, whatever else is needed, or (B) A similar tab on faforever.com.

Bottom line, if you make it easy and intuitive to report bugs, we will do it. It is currently not easy or intuitive (or it's not intuitive that it's easy, which is basically the same thing).

It sounds like in IceDreamer's post above ("they'll be here soon!") something like this is coming, which I'm very excited for.

Again, just want to make sure you know we're all on the same team here.

Statistics: Posted by codepants — 30 Jan 2015, 04:35


]]>
2015-01-29T11:07:43+02:00 2015-01-29T11:07:43+02:00 /viewtopic.php?t=9326&p=92503#p92503 <![CDATA[Re: Suggestion for next patch]]>
SilverKronos wrote:
I just posted this thread to suggest that next patch will be only bug fix, not that i don't like UI improvement and more stat but if these improvements bring more bugs you will need more time too fix it.

i didn't wanted this thread to be a whinning thread to the dev team (is that understandable ?). sry guys :(

For my part, i always post my logs when i detect a bug or a crash ;)


The past releases have actually been bugfix oriented. Minor UI improvements have made it in, but mainly as a side effect of cleanup work on the code.

A lot of work is piling up in the form of un-merged pull requests for the balance patch; precisely because we're focusing on bug fixes: https://github.com/FAForever/fa/pulls?q=is%3Aopen+is%3Apr+milestone%3A%223642+-+Balance+Patch%22.

This means there are 13 units-of-work ready for the balance patch (3642), which are just waiting for us to release the next bugfix patch (3641) before going in.

Statistics: Posted by Sheeo — 29 Jan 2015, 11:07


]]>
2015-01-29T11:08:45+02:00 2015-01-29T11:03:36+02:00 /viewtopic.php?t=9326&p=92502#p92502 <![CDATA[Re: Suggestion for next patch]]>
Xinnony wrote:
Me too and my friend, have a freeze some time, never have this problem before the patch.


There are at least two sim crash/freeze bugs that I know of from < 3634:

https://github.com/FAForever/fa/issues/68 - still unknown

https://github.com/FAForever/fa/issues/2 - cyclic assist which we have a partial fix for.

I'm certain there are others, and I definitely remember experiencing core dump crashes before 3632. I never did get them reported though.

Statistics: Posted by Sheeo — 29 Jan 2015, 11:03


]]>
2015-01-29T11:09:19+02:00 2015-01-29T10:41:37+02:00 /viewtopic.php?t=9326&p=92501#p92501 <![CDATA[Re: Suggestion for next patch]]>
codepants wrote:
I like what you have to say, Sheeo.

I also want to note that since... recently... I've seen an increase both in frequency of patches and in number of bugs.

Maybe if the frequency decreased -- eg, if we tested more before releasing patches -- so the number of bugs would go down? It just seems like we release patches for patches that make more problems, which we release patches for, and so on... I know not every bug or crash is fixable... but in the past few months something changed about the way the game was updated, and it seems like it should change back.


This is an interesting point. Indeed the frequency of patches has increased. The absolute frequency of bugs has likely increased as well, but given how much more code is being churned, I think we have an overall lower relative frequency of bugs.

Check out the development history in terms of commits (Atomic sets of changes that we test indidividually/in groups called pull requests) here: https://github.com/FAForever/fa/graphs/contributors. You'll notice how mountains have appeared since october.

We're listening to this point as well, and the plan is indeed to have publicly playable release candidates ahead of time. Since we're in the midst of reworking how mods work relating to dependencies and loading, we haven't been able to do this as easily as we'd like. With the next release of the python lobby this will be really easy for us, though.


On a final note, regarding the crash bugs -- they're pretty hard to find as they only occur in limited situations, and we're unfortunately dealing with issues in the engine code, making them even harder to debug and fix. The 3640 release was all about fixing a particularly nasty one, and we're left with another one that appears to happen when things crash on ACU's.

All other crash bugs we're getting reports of are not actually new bugs introduced in 3639 -- a release that overall fixes a lot of issues (Changelog here: https://github.com/FAForever/fa/blob/develop/changelog.md#changelog-for-patch-3639) but has only introduced this one obscure crash bug and another one that was fixed within a day with 3640.

We're not counting the lobby here, of course, which has never really seen a remotely bug-free release.

Statistics: Posted by Sheeo — 29 Jan 2015, 10:41


]]>
2015-01-28T15:13:40+02:00 2015-01-28T15:13:40+02:00 /viewtopic.php?t=9326&p=92416#p92416 <![CDATA[Re: Suggestion for next patch]]>
sisso wrote:
Gorton wrote:To get things fixed once they're broken, you need to report them.
Not many people are doing that.
Using a "patch test mod" would be ideal, but past experience has shown that people are less than enthusiastic to test patches (even though they've asked for such a mod to test).


We need to think something to improve this, convert a complain user into a extra hand. I think something like:

1) Release candidates should be very visible in lobby, forum and site. (I don't know where this info is today, forum?)

2) Any release candidate should be played easy (I don't know how, today is a mod?)

3) Should be easy to report, a script, auto crash report, a annex a file from desktop. (I don't know how, I read about logs :P)

4) Any complain about bugs from last patch should be added in a list of testers and query before each release what they think. We have 2 wins here, or he will be fell guilt and test before answer or be quiet next time :P

Or I am talking bullshit?


All four of those things are things the devs want to do, but they require changes to the client itself. So they'll be here soon!

Statistics: Posted by IceDreamer — 28 Jan 2015, 15:13


]]>
2015-01-28T14:50:44+02:00 2015-01-28T14:50:44+02:00 /viewtopic.php?t=9326&p=92412#p92412 <![CDATA[Re: Suggestion for next patch]]>
Gorton wrote:
To get things fixed once they're broken, you need to report them.
Not many people are doing that.
Using a "patch test mod" would be ideal, but past experience has shown that people are less than enthusiastic to test patches (even though they've asked for such a mod to test).


We need to think something to improve this, convert a complain user into a extra hand. I think something like:

1) Release candidates should be very visible in lobby, forum and site. (I don't know where this info is today, forum?)

2) Any release candidate should be played easy (I don't know how, today is a mod?)

3) Should be easy to report, a script, auto crash report, a annex a file from desktop. (I don't know how, I read about logs :P)

4) Any complain about bugs from last patch should be added in a list of testers and query before each release what they think. We have 2 wins here, or he will be fell guilt and test before answer or be quiet next time :P

Or I am talking bullshit?

Statistics: Posted by sisso — 28 Jan 2015, 14:50


]]>
2015-01-28T14:42:13+02:00 2015-01-28T14:42:13+02:00 /viewtopic.php?t=9326&p=92410#p92410 <![CDATA[Re: Suggestion for next patch]]>
i didn't wanted this thread to be a whinning thread to the dev team (is that understandable ?). sry guys :(

For my part, i always post my logs when i detect a bug or a crash ;)

Statistics: Posted by SilverKronos — 28 Jan 2015, 14:42


]]>
2015-01-28T07:59:18+02:00 2015-01-28T07:59:18+02:00 /viewtopic.php?t=9326&p=92400#p92400 <![CDATA[Re: Suggestion for next patch]]> Whatever the reason, it's broken.
To get things fixed once they're broken, you need to report them.

Not many people are doing that.

Using a "patch test mod" would be ideal, but past experience has shown that people are less than enthusiastic to test patches (even though they've asked for such a mod to test).

Remember that a very limited number of people are working.
Also remember that it's their free time they are giving up - and by getting the thousands of players to play the game will show problems much faster than testing. They could of course spend a week making something and then 1 month bugtesting, but that would be a waste of time.
Remember that there are tens of thousands of people who play - and a very small amount have complaints. It's not as widespread as you think.

Statistics: Posted by Gorton — 28 Jan 2015, 07:59


]]>
2015-01-28T06:52:19+02:00 2015-01-28T06:52:19+02:00 /viewtopic.php?t=9326&p=92395#p92395 <![CDATA[Re: Suggestion for next patch]]>
Another problem we've had is the sheer amount of code being changed in the lobby under the hood to make it a bit easier for developers to keep on top going forwards. It was always going to get worse before it got better, though we had hoped not this much. One unforeseen circumstance was lobby mods people use hooking into code that was made defunct, which caused some issues. Another thing noticed is that, given the increased scrutiny from the community post-patch, a lot of bugs in the lobby are being reported as though they are new, when in fact it turns out they are not, and were there previously.

Everyone's been hard at work fixing all the regressions, improving the code, and improving the visuals behind the lobby, as well as the server backend, over the past few weeks so that the next patch (Which is coming soon-ish!) will be focused on getting the lobby and server working really, really smoothly, as well as a handful of fixes to the actual game (For example, Cybran engineer build graphics adjustments which were suggested, reclaim per second being shown in the panel, etc).

As far as the hard crashes people are reporting... I don't know enough about it personally, but Sheeo's trying to figure out the cause. It's a tricky one imo, because though I've looked at the game logs, nothing really gets logged when the sim outright crashes, which makes tracking down the cause really hard.

Statistics: Posted by IceDreamer — 28 Jan 2015, 06:52


]]>
2015-01-28T05:11:16+02:00 2015-01-28T05:11:16+02:00 /viewtopic.php?t=9326&p=92391#p92391 <![CDATA[Re: Suggestion for next patch]]>
codepants wrote:
Gorton wrote:
cannot fix them without reports of what's happening


So you're saying that when you do pre-release testing, everything works perfectly without any new bugs?

Out of curiosity, how much testing do you do? Not looking to judge or criticize, just curious.


Lol, exactly !
The Lobby have lot of bug too, it is impossible to miss.
There just has not been a test.

Statistics: Posted by Xinnony — 28 Jan 2015, 05:11


]]>
2015-01-28T03:56:42+02:00 2015-01-28T03:56:42+02:00 /viewtopic.php?t=9326&p=92387#p92387 <![CDATA[Re: Suggestion for next patch]]>
Gorton wrote:
cannot fix them without reports of what's happening


So you're saying that when you do pre-release testing, everything works perfectly without any new bugs?

Out of curiosity, how much testing do you do? Not looking to judge or criticize, just curious.

Statistics: Posted by codepants — 28 Jan 2015, 03:56


]]>
2015-01-27T21:36:45+02:00 2015-01-27T21:36:45+02:00 /viewtopic.php?t=9326&p=92365#p92365 <![CDATA[Re: Suggestion for next patch]]>

cannot fix them without reports of what's happening

Statistics: Posted by Gorton — 27 Jan 2015, 21:36


]]>
2015-01-27T04:41:03+02:00 2015-01-27T04:41:03+02:00 /viewtopic.php?t=9326&p=92309#p92309 <![CDATA[Re: Suggestion for next patch]]>
SilverKronos wrote:
hi, since i see a lot of people crashing these days, here are my thoughts
the more patch we have, the more bug we have....

i suggested that next patch will be only correction of bug like the game crashing.
Those bug are really game breaker ...

cheers

SilverKronos


Me too and my friend, have a freeze some time, never have this problem before the patch.

Statistics: Posted by Xinnony — 27 Jan 2015, 04:41


]]>
2015-01-27T04:29:19+02:00 2015-01-27T04:29:19+02:00 /viewtopic.php?t=9326&p=92305#p92305 <![CDATA[Re: Suggestion for next patch]]>
ckitching wrote:
codepants wrote:but in the past few months something changed about the way the game was updated, and it seems like it should change back.


Correlation does not imply causation.


Ooh, what a grand statement. It must not be the patches, then, it must be the wind.

I don't care about the cause, I want a playable game. I would even stand for *no* patches if that meant everything worked as well as it did a year ago.

I love everything the team is doing. I love the game. I'm just saying, I'd be okay with slowing down the rate of patches if that meant the game was more playable.

(note the "if," Mr. Grand Statements)

Statistics: Posted by codepants — 27 Jan 2015, 04:29


]]>